The Notion of Null Hypothesis in Generative Grammar

다국어 초록 (Multilingual Abstract)

This paper claims that given the exhaustive list of interpretive effects of Merge, as argued in this paper, the formal operation of Merge may be null hypothesis, which leads to new linguistic explanations, crucially supporting non-determinism (Yang 2013, 2014). Chomsky (2008) states: Merge can apply freely, yielding expressions interpreted at the interface in many different kinds of ways. The free Merge theory as claimed by Chomsky (2008) above implies that the Merge operation is free and optional unless the derivation comes to be constrained by some condition like the [+marked] feature, as discussed with respect to non-determinism (Yang 2013, 2014). The free and optional Merge is unmarked unless it is marked by being obligatory due to the [+marked] feature as mentioned above. Chomsky (2008) claims that every Merge, especially the unmarked Merge, always induces an interpretive effect. In fact, we may assume that even the marked Merge, which has been often exceptionally claimed to induce no interpretive effect, can be assumed to induce an interpretive effect called the null interpretive effect, or simply null effect. Hence, we accept Chomsky's claim literally that every Merge induces an interpretive effect, and develop the theory of grammar crucially based on the general notion of interpretive effect of Merge. The overt exact operation of Merge for each interpretive effect is rarely distinctively defined in contrast to the general (semantic and/or functional) notion of interpretive effects like topic, focus, etc., so that the overt operation of Merge itself can be assumed to be null hypothesis, obviously even without involving the edge feature effects (Chomsky 2008), while the essential distinctive effects of Merge should be the semantic or functional interpretive effect itself (Yang 2014). By way of motivating the null hypothesis theory for Merge as characterized above, I will discuss significant theoretical implications of the null hypothesis of Merge with respect to new linguistic explanations, speculating how far the null hypothesis theory is applicable in our grammar.